Nonreciprocal routing induced by chirality in an atom-dimer waveguide-QED system

TL;DR

The work studies directional control of single-photon transport in a four-port waveguide-QED device composed of two dipole-coupled TLAs chirally coupled to two waveguides. The authors derive exact single-photon scattering amplitudes with the real-space method and analyze both Markovian and non-Markovian dynamics by incorporating the propagation time $\tau$, showing that chirality quantified by $G=\gamma_2/\gamma_1$ induces nonreciprocity. They demonstrate regimes where perfect directional routing is achieved by tuning the dipole coupling $\xi$ together with $G$, and show that non-Markovian effects can enhance nonreciprocity and relax perfect chirality requirements. The results point to a feasible path toward nonreciprocal, chiral quantum routing in waveguide-QED platforms and suggest experimental implementations in superconducting circuits and photonic systems.

Abstract

The implementation of quantum routers is an important and desired task in quantum information science, since quantum routers are important components of quantum networks. Here, we propose a scheme for implementing single-photon routers in a waveguide-QED system, which consists of two coupled two-level atoms coupled to two waveguides to form a four-port quantum device. We obtain the exact analytical expressions of the single-photon scattering amplitudes using the real-space method. By taking the propagating time of photons between two coupling points into account or not, we consider the system working in the Markovian and non-Markovian regimes, respectively. In addition, we introduce the chiral coupling, which breaks the symmetry of the waveguide model, to manipulate the transmission of single photons. We find that when the system works in the non-Markovian regime, the single photon can be transmitted on demand by adjusting the asymmetry coefficient. More interestingly, the complete single-photon routing in this device does not require an ideal chiral coupling, loosening the photon transport conditions. This work will motivate the studies concerning the nonreciprocal and chiral quantum devices in the waveguide-QED platform.

Figures (7)

• Figure 1: Schematic of the four-port waveguide-QED system. Two independent waveguides $a$ and $b$ interact with two TLAs, which are coupled with each other via the dipole interaction with the coupling strength $\xi$. The $i$-th $(i=1,2)$ TLA couples to waveguides $j$ at the coupling point $x_{j_{i}}$. Here we take $x_{j_{1}}=0$ and $x_{j_{2}}=L$$(j = a, b)$.
• Figure 2: Transmission coefficients (a, b) $T_{j}^{(M)}$$\left(j=a,b\right)$ for the photon input from port 1 and (c, d) $\widetilde{T}_{j}^{(M)}$ for the photon input from port 2 vs the detuning $\delta/\gamma_{1}$ and the phase shift $\theta/\pi$ in the Markovian regime. The white dashed curve indicates $\widetilde{T}_{a}^{(M)}=0.99$. Other parameters used are $G=2.38$ and $\xi/\gamma_{1}\!=\!0.5$.
• Figure 3: (a) Contrast ratio $\left|I_{a}^{\left(M\right)}\right|$ vs $G$ and $\xi/\gamma_{1}$. The black solid curve indicates $\left|I_{a}^{\left(M\right)}\right|=0.99$. (b) The profiles of (a) at $\xi/\gamma_{1}=0$ (black dotted curve), $\xi/\gamma_{1}=1$ (blue solid curve), and $\xi/\gamma_{1}=2$ (red dashed curve). (c) The profiles of (a) at $G=0$ (black dotted curve), $G=1$ (blue solid curve), and $G=2$ (red dashed curve). Other common parameters used are $\theta/\pi=0.1,~\omega_{e}/\gamma_{1}=100$, and $\varepsilon/\gamma_{1}=103.$
• Figure 4: (a, c) Transmission coefficients $T_{j}^{M} (j=a,b)$ as functions of the chiral parameter $G$ and dipole coupling strength $\xi/\gamma_{1}$ for the photon incident from port 1. The white and black dashed curves indicate $T_{j}^{\left(M\right)}=0.9$ and 0.99, respectively. The curves in (b, d) show the profiles of (a) and (c) at given values of $G$ and $\xi$. The other common parameters used are $\theta/\pi=0.1$, $\varepsilon/\gamma_{1}=103.4$, and $\omega_{e}/\gamma_{1}=100.$
• Figure 5: Transmission coefficients (a, b) $T_{j}^{(N\!M)}$$\left(j=a,b\right)$ for the photon input from port 1 and (c, d) $\widetilde{T}_{j}^{(N\!M)}$ for the photon input from port 2 vs the detuning $\delta/\gamma_{1}$ and the scaled propagating time $\gamma_{1}\tau$. In all panels, the parameters are $G\!=\!2.38,~\xi/\gamma_{1}\!=\!0.5$, and $\theta/\pi\!=\!0$.